Snapshot
Continue to About ↓- What it does
- Reflexio is a learning layer for AI agents that captures user corrections and failed tool calls, turns them into reusable behavioral rules, and retrieves those rules before the agent's next response so it avoids repeating the same mistake.
More detail
- It does not retrain the underlying model; it supplies learned context through SDKs, REST, CLI, or a coding-agent skill.
- Best for
- Teams running production AI agents with tool calls and multi-step workflows who want to reduce repeat execution errors and govern what the agent learns.
More detail
- Also suited to teams with compliance constraints that need BYOC or self-hosted deployment.
- Pricing
- Free plan at $0/mo includes 100 learnings and 1,000 search requests per month with a 30-day Pro trial.
More detail
- Pro is $299/mo for 10K learnings and 100K search requests, with overage at 0.3¢ per search request, 2¢ per playbook learning, and 1¢ per profile learning.
- BYOC pricing is undisclosed ('Contact us')
- Main limitation
- Teams wanting a drop-in memory store or model fine-tuning, and teams unwilling to actively review and reject learned rules through the portal.
More detail
- Very low-volume users may find the free tier limits restrictive.
- Record
- Living and monitored
- Reviewed
- Evidence
- Public web research
- Machine access
- Structured profile ↗
About
- Reflexio is a learning layer for AI agents that captures user corrections and failed tool calls, turns them into reusable behavioral rules, and retrieves those rules before the agent's next response so it avoids repeating the same mistake.
- It does not retrain the underlying model; it supplies learned context through SDKs, REST, CLI, or a coding-agent skill.
What it helps with
- Repeat mistakes across conversations because user corrections disappear when the conversation ends
- Missed edge cases in a single turn, such as an agent refunding one unfamiliar charge but missing another
- Redundant reasoning and tool calls that inflate token usage and latency
- Execution failures from repeated tool errors and invalid execution paths
- Uncontrolled agent learning without a way to inspect, edit, or reject learned rules
Capabilities
A practical view of what is available and where limitations apply.
| Offering | Available | Detail |
|---|---|---|
| Data ingestion | Yes | Connects to an instrumented tracing platform or publishes directly from the application; ingests execution paths, tool calls, and user corrections. |
| Intelligence engine | Yes | Extracts recurring failures and synthesizes actionable execution rules from failure logs. |
| Runtime orchestration | Yes | Retrieves relevant personalized rules before each execution and feeds real-response feedback back into the loop. |
| Governance portal | Yes | Inspect, edit, or reject learned rules; view evidence and evaluation results; export, delete, or search learned intelligence. |
| Evaluation | Yes | Measure how learned context changes agent responses. |
| Model fine-tuning | No | Reflexio supplies learned context but does not change model weights. |
| Memory store | No | Explicitly positioned as behavior learning, not conversation-history memory. |
| BYOC self-hosting | Yes | Deploy inside your own AWS, GCP, or Azure VPC with your own LLM and storage; pricing undisclosed. |
| Open-source local deployment | Yes | Apache-2.0 package with local FastAPI backend, SQLite storage, and no authentication by default. |
| Independent benchmark verification | Unknown | Vendor-published benchmark claims (e.g., >36% execution failure reduction, >50% token reduction) were not independently replicated in the sources reviewed. |
Bot trust
How confidently AI systems can verify this profile from public evidence.
—of 100
ChatGPT—/100
Claude—/100
This profile has not completed a Bot Score assessment yet.
Evidence
Where this record came from, when it was reviewed, and how confidently it can be used.
- Verification
- Scribble researched these facts from the linked public sources.
- The owner has not manually confirmed every field.
- Last reviewed
- Primary source
- reflexio.ai
Technical profile details
Record ID
1f3c54e1-2930-4da4-a6f4-1462e82a538eCanonical URLhttps://profiles.scribble.network/listing/reflexio-1f3c54e1